| With the rapid development of socio-economic and scientific technology,reservoir management authorities have put forward a series of recommendations on the accuracy of runoff prediction and the efficiency of reservoir scheduling and operation.has become more and more demanding.In reservoir dispatching,risk dispatching based on runoff forecast information has become a demand.Scientific and rational reservoir dispatching helps to improve water resource efficiency,and now,considering runoff forecast information has become a requirement to improve reservoir dispatching.important non-engineering means of achieving benefits.The impact of deterministic runoff processes on reservoir scheduling is deterministic,while the application of deterministic runoff processes in reservoir scheduling is relatively simple and relatively fixed dispatch model.In practice,however,dispatchers are always interested in applying forecast runoff to guide reservoir scheduling,so this paper will use a numerical Rainfall forecast information to improve the accuracy of medium-term runoff forecast information and the use of the reporting child model to assist dispatchers in decision making.The research and conclusions of this paper are as follows.In this paper,the probabilistic runoff prediction model is firstly developed based on the available information.The probabilistic runoff prediction model is based on Bayesian theory,and the measured runoff information is combined with the GFS deterministic runoff prediction using the full probability formula.for integration.A priori distributions are used in the model to reflect the uncertainty of measured runoff,and likelihood distribution functions are used to reflect model parameters,model structure,etc.aspect uncertainty in runoff forecasting,using the Bayesian theory of the posterior distribution probability density function to represent the uncertainty in runoff forecasting,using the posterior The mathematical expectation of the distribution probability density function is used as the deterministic forecast output of the model,with a 90% confidence interval as the model’s Forecast interval output.Then,the reservoir scheduling decision gradually deviates from the actual optimal decision due to the influence of runoff uncertainty.On the basis of the runoff probability forecast,the coupled model of the announcer boy model and the reservoir scheduling model is developed,and the model is used to solve the runoff The effect of uncertainty on reservoir scheduling decisions.In this model decision makers make scheduling decisions for reservoirs through revenue setting conditions and consider the risk under reservoir scheduling decisions Change scenario.Based on the setting of benefits,make the best decision consistent with the scope of risk management.Finally,consider the effect of uncertainty in runoff forecast information on the effectiveness of reservoir dispatch,and consider the change in runoff forecast information with forecast duration The reservoir scheduling model based on the prognosticator model is developed by merging the runoff forecast information with the demand distribution in the prognosticator model.The model focuses on the study of the impact of runoff uncertainty information on the formulation of efficient and reliable reservoir dispatching decisions,and is based on the probabilistic prediction of runoff.The forecast interval is the input interval and the decision interval is derived for certain risk conditions.The analysis of the changes in the decision interval is used to provide decision makers with the basis for risk scheduling decisions.(1)The Bayesian theory-based runoff probability forecast is first analyzed for anomalous data in the runoff probability forecast results,resulting in anomalous The main reason for the data is the low degree to which the measured runoff process satisfies the normal linearity assumptions in the model.The input data are then transformed to satisfy the normal linearity assumptions to obtain reliable forecast results with the required accuracy.When the uncertainty of the deterministic prediction of the runoff process is determined,the degree of agreement of the measured runoff process with the conditions of the model determines the runoff accuracy.Forecast values as well as the accuracy of the forecast interval.(2)Reservoir risk dispatching using the announcing child model,which is the first time to combine the announcing child model with reservoir dispatching model.The model first compares the simulated reservoir runoff with the measured runoff and the probabilistic predicted runoff to obtain the runoff process of the model.better than the mean of the probabilistic forecasts and more consistent with the measured runoff process.Then the revenue constraint is set under different probabilistic expectations.The risk variation analysis of reservoir scheduling decisions under different revenue constraints provides a theoretical basis for decision makers’ reservoir scheduling decisions.(3)In the case of risk setting,the change of the runoff forecast interval will directly affect the change of the decision feasible interval.The accuracy of the runoff forecasting interval directly affects the accuracy of the model.At a certain forecast interval,an increase in risk will lead to an increase in the width of the feasible interval and ultimately to an increase in the generation capacity of the dispatch simulation.For the use of decision feasible intervals in the model,decision makers can set their own level of risk tolerance based on their risk appetite,the selection interval is then predicted based on the range of feasible intervals,and finally model simulations are used to derive the maximum return under the risk tolerance. |
